We have studied the Doushantuo phosphorite fossils for more than a decade, and we disagree with the interpretation of Li et al. that some of the submillimeter microfossils were parenchymella larvae and amphiblastula embryos of sponges (1).

The putative parenchymella larvae of sponges, with a shoe-shaped morphology and dense peripheral flagella, as described by Li et al., are actually acanthomorphic acritarchs. Acritarchs are organic-walled microfossils that cannot be placed confidently into any existing classification categories, but most are comparable to remains of cysts of planktic eukaryotic algae (8). The acanthomorphic acritarchs (an acritarch group with distinct peripheral processes) are common and abundant in Neoproterozoic and Paleozoic deposits. Each individual acanthomorphic acritarch consists of an ovoid body (vesicle) surrounded by processes (Fig. 1, A and B). The processes of different taxa are varied in shape and size: mammillate, spine-form, hair-like, or flagellum-like. The specimen illustrated by Li et al., figure 2E of their report (1), is a deformed individual, with a shrunk vesicle and hair-like (or spine-like) processes. This specimen is morphologically comparable to those of the acritarch genus Ericiasphaera, which was set by Vidal in 1990 (9) and was redescribed to apply the large sphaeroidal acritarchs bearing numerous regularly arranged, solid, flexible, or rigid processes (6). Three species from the Doushantuo Formation have been described and assigned to this genus (6). We have observed many acanthomorphic acritarch specimens in the Doushantuo phosphorites (4, 6, 10), which display a wide range of morphological variation resulting from taphonomic changes: from a relatively intact specimen (Fig. 1, A and B), to the shrunk forms at different degradation states (Fig. 1, C and D). Some of these specimens are similar to those illustrated by Li et al. (figure 1C in the report).

The granular and variable structures within the phosphorites have been interpreted as sclerocytes (sponge skeleton-setting cells) and amoebocytes, respectively, by Li et al. (1). However, as the figures in their report show, these structures have no distinct boundary membranes. Instead, they consist of amorphous kerogen. We have seen such structures in the vesicles of acritarchs (Fig. 1B) or in fossil remains (Fig. 1, E and F). These structures may not be original biological structures, but amorphous kerogenous forms resulting from organic degradation and diagenetic processes.

The structure interpreted as the sponge's amphiblastula embryo by Li et al. is actually a part of the alga thallus. Similar structures have often been observed from the Doushantuo phosphorites; they have been assigned to Wengania globosa, Zhang, with global thallus consisting of rectangular, polygonal, and irregular cells, and possible affinity of rhodophytes (Fig. 2A). This form of algae was first found and described in 1989 from Wengan Phosphate Mine by one of us (2) and was redescribed subsequently in several papers (3, 6). The so-called amphiblastula shown by Li et al. [figure 2F in (1)] are the same as the thallus of W. globosa. The tissues of the fossil algae are composed of regularly arranged cells, which display clear cell wall and condensed cytoplast (Fig. 2A and C). The tissue of modern rhodophytes, in a sectional view, shows morphological features similar to those of the fossils (Fig. 2B). The figure provided by Li et al. for the amphiblastula embryo [figure 2F in (1)] shows regularly arranged rectangular cells, cell-walls, and condensed cytoplast. Thus, the specimen of "amphiblastula" is an algal thallus (a broken thallus) rather than a sponge embryo.

The tiny spine- or needle-formed structures interpreted as monaxonal spicules of sponges by Li et al. are questionable. These structures in the phosphorite rocks could be multiple in origin. They could be detached broken spines of collapsed acritarchs (Fig. 3, A and B).

Our criticism does not mean that the Doushantuo Formation has only preserved remains of algae. On the contrary, we have pointed out that, among the discovered Doushantuo fossils, "a few forms exhibit certain features of metazoan tissue" (3). In addition, triact spicules have been found in the cherts of the Doushantuo Formation in Yangtze Gorge area (11), and we have found that the Doushantuo phosphorites preserved early-staged embryos of possible bilaterally symmetrical animals (5).

We expect more convincing evidences for early animals and plants to be found from the Doushantuoan and other contemporaneous or even older deposits in the near future.

Yun Zhang
College of Life Science,
Beijing (Peking) University,
Beijing, 100871, China
E-mail: yunzhang{at}pku.edu.cn
Xunlai Yuan
Leiming Yin
Nanjing Institute of Geology and Paleontology,
Chinese Academy,
Nanjing 210008, China
E-mail: algae{at}pub.nj.jsinfo.net

REFERENCES

1. C.-W. Li, J. Y. Chen, T. E. Hua, Science 279, 879 (1998) [Abstract/Free Full Text] .
2. Y. Zhang, Lethaia 22, 113 (1989) [ISI], in Proc. 30th Internatl. Geol. Congr. H. Z. Wang et al., Eds. (VSP, Zeist, The Netherlands, 1997), vol. 1, pp. 187-199.
3. Y. Zhang and X. L. Yuan, Lethaia 25, 1 (1992) ; Y. Zhang and X. L. Yuan, Science in China Series C 39, 28 (1996) .
4. X. L. Yuan, Q. F. Wang, Y. Zhang, Acta Micropaleontologica Sinica 10, 409 (1993) .
5. S. H. Xiao, Y. Zhang, A. H. Knoll, Nature 391, 553 (1998) [CrossRef] .
6. Y. Zhang, L. M. Yin, S. H. Xiao, A. H. Knoll, Paleontol. Soc. Mem. No. 50 (1998), pp. 1-52.
7. S. Golubic and E. S. Barghoorn, in Fossil Algae E. Flugel, Ed. (Springer-Verlag, New York, 1977), pp. 1-14; S. Golubic and Y. Zhang, Hydrobiologia 123, 193 (1985) [CrossRef] [ISI].
8. C. V. Mendelson, in Fossil Prokaryotes and Protists, J. H. Lipps Ed. (Blackwell, Boston, 1993), pp. 77-104.
9. G. Vidal, Paleontology 33, 287 (1990) .
10. X. L. Yuan and H. J. Hofmann, Alcheringa, in press.
11. Z. Zhao et al., Biostratigraphy of the Yangtze Gorge Area, Sinian, part 1 (Geological Publishing House, Beijing, 1985), p. 143; The Sinian System of Hubei Province (China Univ. Geosciences Press, Wuhan, China, 1988), p. 205.

Response: In our report (1), we described microscopic sponges and metazoan embryos from the 580-million-year-old Doushantuo phosphorites at Wengan, South China. Recently, Xiao et al. described multicellular algae and animal embryos from the same phosphorites (2). These discoveries not only offered a glimpse of familiar-looking animals before the Cambrian explosion, but also have implications for the study of early animal evolution (3). Recently, after examining more than 3000 embryo fossils, we are now able to reconstruct the sequential embryogenesis of the Wengan sponge. Zhang and his colleagues have done the pioneering work on describing well-preserved acritarchs and multicellular algae from the Wengan phosphorites (4-7). However, they have misidentified some of the abundant globular metazoan embryos as phytoplanktonic organisms (7, 8), and still identify an array of organisms as multicellular algae or as ambiguously defined acritarchs.

In their comment, Zhang et al. propose that spicules of the sponge that we described could be aggregates of detached broken spines of collapsed acritarchs. We disagree and would like to present further evidence (Fig. 1) in support of the identifications we made in our report. We have examined three specimens of this type of mushroom-shaped sponges. One specimen (Fig. 1) is a juvenile, and consists of a main crown, a slender stalk, and a flattened holdfast. It has been cracked into two halves during the late diagenesis, and the cracking space has been filled with diagenetic silica. That the spicules occur only in the body of the sponge, not anywhere along the crack or in surrounding matrix, indicates that they are biogenic in nature, not diagenetic products. The most important features of this sponge specimen are the preservation of the surface ornaments and the abundant interior spicules. The surface of this organism is ornamented with tiny knobby structures, which are regularly spaced and triangular in shape (Fig. 1A). The spicules occur randomly within the distinct body, and all spicules are monaxial, with two tapering ends (Fig. 1B). The axial canals could be observed in well-preserved spicules (Fig. 1, C and D). These characteristics fit well with the criteria of the sponge's spicules. Furthermore, we have illustrated two sclerocytes that attach to one end of each forming spicules. Both the sclerocytes are bounded by a distinct plasma membrane, which was marked as "pm" in figure 1F in our report, and which indicates a distinct boundary membrane.

In their comment, Zhang et al. refer to some organic remains with peripheral processes as the products of "collapsed acritarchs resulting from taphonomic changes," and stated that these organic remains are comparable to the parenchymella larva described by us. We have quite different views on this issue. First, the specimens shown by Zhang et al. are evidently different from our specimen. The peripheral flagella of our specimen not only taper to their distal ends, but also have a flexible outline, while the peripheral processes of their specimens are uniform in diameter, with broadly rounded distal ends. We aim to be cautious in describing the larva. Our tentative identification is mainly based on the shoe-shaped morphology and dense peripheral flagella. However, there are many kinds of organisms that could share these characteristics, including protists and planula larvae of coelenterates, so we called it "possible parenchymella larva" or "parenchymella-type larva" consistently.

Second, the purported "collapsed acritachs" with peripheral processes are not only different from our "possible parenchymella larva" specimen, but also appear to be filamentous bacteria growing on organic remains. We have examined tens of this type of specimen, which are comparable to the specimens described as collapsed acritarchs, and we conclude that these filamentous bacteria (Fig. 2, A-D) are quite different from both the peripheral processes of the acritarchs and the flagella of the parenchymella larvae. These filaments are uniform, with a diameter of ~0.75 µm through their long axes, length varying from 20 µm to 93 µm, and distal ends that are broadly rounded. This kind of association is a common phenomenon in extant aquatic environments. A reasonable explanation is that they were fossilized in situ, keeping their original shape through the taphonomic process. That is the beauty of the Wengan phosphorites in which almost all the organisms, including these bacteria, are kept intact and preserved in three dimensions. The acritarch (Gr. "of uncertain origin"), a loosely defined group, has become a grab bag containing a variety of unicellular microfossils. Although some are the right size and shape to be the coats of cysts similar to those made by certain extant protists, acritarchs should be much more polyphyletic than what palaeontologists generally conjectured.

We have assigned a specimen as a possible amphiblastula larva based on the similarity between our specimen and the amphiblastula larva of an extant sponge (Grantia compressa) (9, p. 77) in morphology and arrangement of component cells. Zhang et al. tend to interpret our specimen as an algal thallus (Wengania globosa), but there are fundamental differences between our "possible amphiblastula larva" and the specimen of W. globosa shown by Zhang et al. By comparing the schematic drawings (Fig. 3) that reveal the anatomy of our "possible amphiblastula larva" and W. globosa, one can observe the distinct differences in size, morphology, and arrangement of the component cells between them. Furthermore, the description of W. globosa in their comment seems inconsistent with other publications of theirs. Although the very same figure of this taxon shown in the comment by Zhang et al. was apparently used earlier by them (5), there seem to be discrepancies between the associated legends concerning the species names and the size of the thallus. In that paper (5), this organism was referred to as W. rotatoria and the size of the thallus was stated as 393 × 330 µm not 86 × 73 µm as stated in the comment. The genus Wengania was also loosely defined. Zhang and Yuan erected a new genus Cerionopora in 1992 (5) and briefly described the diagnostic difference between it and Wengania, but in 1993, they illustrated the same figure [plate I, figure 3 in (6)] under the name W. globosa without apparent explanation.

The discoveries of microscopic animals and embryo fossils from Wengan phosphorites (1, 2) have created some challenges for scientists to tackle. Further research and evaluation is needed to clarify the ultrastructure of the phosphatized fossils in order to reveal the scope and significance of the Wengan biota.

Chia-Wei Li
Department of Life Science,
National Tsing Hua University,
Hsinchu, Taiwan
E-mail: lslcw{at}life.nthu.edu.tw
Jun-Yuan Chen
Nanjing Institute of Geology
and Palaeontology,
Academia Sinica, Nanjing210008, China
E-mail: chenjy{at}jlonline.com
Tzu-En Hua
Department of Life Science,
National Tsing Hua University,
Hsinchu, Taiwan
E-mail: d865402{at}oz.nthu.edu.tw

REFERENCES

1. C.-W. Li, J.-Y. Chen, T.-E. Hua, Science 279, 879 (1998) .
2. S.-X. Xiao, Y. Zhang, A. H. Knoll, Nature 391, 553 (1998) .
3. S. Bengtson, ibid., p. 529.
4. Y. Zhang, Lethaia 22, 113 (1989) .
5. Y. Zhang and X.-L. Yuan, ibid. 25, 1 (1992).
6. X.-L. Yuan, Q.-F. Wang, Y. Zhang, Acta Micropalaeontol. Sinica 10, 409 (1993) .
7. L.-M. Yin and Y.-S., Yue, Chinese J. Botany 5, 168 (1993).
8. Y.-S. Xue, T.-F. Tang, C.-L, Yu, C.-M. Zhou, Acta Palaeontol. Sinica 36, 688 (1995) .
9. L. De Vos, K. Rützler, N. Boury-Esnault, C. Donadey, J. Vacelet, Atlas of Sponge Morphology (Smithsonian Institution, Washington and London, 1991).